6

WHAT THE MIDDLE AGES MADE OF THE WITCH

THUS FAR, AN examination of the relationship between specific themes in medieval culture and their ancient antecedents has revealed some perhaps unexpected contrasts. In the case of ceremonial magic, dependent on the transmission of texts copied by a literate elite, the continuity with the ancient world, and the survival of names and ideas from it, seems to have been greater than has traditionally been considered. On the other hand, in the case of beliefs concerning spectral processions of the night, which have long been thought by scholars to have derived directly from pre-Christian tradition, the lines of transmission seem to be a lot harder to prove than has been presumed. It remains to confront directly the issue of how the Middle Ages dealt with the central theme of this book, the figure of the witch. This chapter will tackle that problem and in doing so propose answers to three more: what difference the establishment of Christianity as the dominant religion of Europe made to attitudes to magic and witchcraft; how seriously witchcraft was treated in the course of the Middle Ages; and how the stereotype of the witch as a practitioner of a satanic religion, which underpinned the early modern witchcraft trials, came to evolve.

The Immediate Impact of Christianity

It was remarked earlier that the Christian religion which underpinned the early modern witch trials combined the whole range of ancient traditions which individually established parts of a context for witch-hunting: Mesopotamian demonology; Persian cosmic dualism; a Graeco-Roman fear of magic as intrinsically impious; Roman images of the evil witch; and the Germanic concept of night-roaming cannibal women. Comments by respected scholars have not been lacking, indeed, to credit the Christian faith with an inherent propensity to encourage the persecution of magicians. Valerie Flint has argued that its institutionalized and monopolistic traits made it automatically into a state religion that demanded tighter control of human dealings with spirits, most of which became evil by definition.¹ Richard Kieckhefer has pointed out that Christianity redefined magic in a totally new way, as the worship of false gods, alias demons.² Michael Bailey has agreed, observing that Christians always posited a more fundamental distinction between religion and magic than that imagined by pagans and Jews.³ All this is correct, but there are two obvious features of the history of magic that provoke counterbalancing reflections. One is that the early European witch trials commenced a thousand years after the triumph of the new religion, raising the problem of why, if its ideology was so well suited to hunt witches, it took so long to do so. The second is that, as discussed earlier in this book, the pagan Roman Empire had proved perfectly capable of enacting a savage code of laws against magicians, based on wholly traditional attitudes, at precisely the same time as it was persecuting Christians with an equal brutality. It is, in fact, that legal and cultural context, of established and intense official hostility towards magic, that provides the reasons for the Christian perspective on the subject. It presented early Christians with an acute problem: that the miracles credited by them to their Messiah and his apostles could look like those promised by, or attributed to, ceremonial magicians. This charge was levied against them by some of their most effective pagan critics, such as Celsus, who wrote the first comprehensive attack on the new religion in the second century. The reply provided by the leading Christian theologian Origen became the standard one: magicians used rites and incantations, but true Christians only the name of Jesus and the words of the Bible, and a reliance on the power of their deity: a formula which plugged directly into the long-established Graeco-Roman distinction between religion and magic.⁴ Almost two hundred years later, Augustine of Hippo worked it up into its enduring form, which persisted through the Middle Ages: that the acts of magicians were accomplished with the aid of demons, whereas the miracles of Christian saints were made possible by the intervention of the one true God.⁵

The polemical position that Christianity established with regard to magic, therefore, was a defensive one formulated to cope with a serious challenge to its own credibility and public image, and set firmly in the context of existing Graeco-Roman attitudes. It also, however, drew on essential traits of its own, one of which was an extreme manifestation of the Mesopotamian (and thus the Hebrew) fear of demons. Even by the traditional standards of the Fertile Crescent, early Christian demonology was uniquely polarized, depending on the acceptance of a cosmic force of pure evil in the universe, and all-pervasive. The driving of demons out of people whom they had possessed and were afflicting was a chief task of Christ himself, his apostles and the early saints, and these evil spirits were their main enemies, as any perusal of the New Testament, Apocrypha and early hagiographies reveals. On the whole, however, these same texts were much more concerned with direct confrontations between Christian champions and demons than between those champions and human servants of the demons. The problem that early Christians had with (pagan or Jewish) magicians focused on them as rival exorcists and healers, and also as objects of official suspicion and condemnation with whom Christians might too readily be conflated, as Celsus actually did conflate them. The ‘magicians’ Simon Magus and Elymas, and itinerant Jewish exorcists of the classic ancient Mesopotamian sort, already feature in the Acts of the Apostles as fools or charlatans, and one of the traits of the Whore of Babylon in the Book of Revelation is pharmakeia, still the standard Greek word for potion-based magic.⁶ These motifs were multiplied and amplified in subsequent early Christian literature, but never add up to make that literature a body of witch-hunting documents. The magicians portrayed are too weak, and easily bested by Christian holy men, to represent dangerous opponents, and there is none of the special association of women with bad magic found in ancient Roman and Jewish sources.⁷ Jesus himself was not interested in magic, and when the apostle Paul condemned it, he did so as a sin on a level with anger and lechery, and not as a lethal crime.⁸ All this forms the background to what happened after the year 312, as Christianity became the dominant religion in the empire and the one professed by most of its emperors (and by all after the year 363). In the fourth century, a series of church councils passed decrees which forbade Christians, and clergy in particular, to have anything to do with ceremonial magic, and included divination as part of that.⁹ Imperial law adapted accordingly during the same century and the next, the big shift in it being to redefine practices that had been normative in ancient paganism, such as divination by official temple personnel and sacrifice as a religious act, as superstitious or magical, and therefore forbidden. When dealing with ceremonial magic, however, and harm caused by magical means, the laws did little more than reinforce what had already been laid down under pagan emperors.¹⁰ This continuity would none the less still mask significant change if the existing laws against ceremonial magic, and all forms of harmful magic, were enforced more rigorously than before.

An impression that this was indeed the case is provided in a series of famous passages by the fourth-century historian Ammianus Marcellinus. He recorded that following a law of Constantius II, in 358, which declared that magicians across the empire were enemies of humanity, anybody who wore an amulet to cure a disease, or passed a tomb after dark, was in danger of denunciation and execution. Being seen near a tomb was fatal because of suspicion that the person concerned was hunting for human body parts for use in spells. This wave of trials was followed by three more, at intervals between 364 and 371, under the brother emperors Valentinian and Valens. Those began by mainly affecting the Roman senatorial class, but expanded over time to target commoners. Whole libraries were burned by their owners for fear they might be thought to contain magical texts. These persecutions affected both Rome itself and the eastern provinces, and torture was freely used to obtain evidence. Ammianus made clear that in most cases the pressure to prosecute came from the top, from emperors leading recently established and insecure dynasties and afraid of conspiracy: the charge of using magic had returned, for the first time in three hundred years, as a weapon in central politics.¹¹ His picture of serious persecution in that century is supported by a text from the 330s, a handbook on astrology, one of the main forms of divination, by Firmicus Maternus. It contains no less than seven examples of horoscopes cast to determine whether the person concerned would be charged with using magic!¹² The sense of a fourth-century society at least at times gripped by a fear both of magic itself and of accusations of using it, is further confirmed by the work of Libanius, a pagan scholar of the mid to late part of the century. Having moved around the eastern provinces, he settled in Antioch, one of the four most important cities of the empire, to become its leading philosopher and orator. His writings contain many reflections on his career, which reveal that the charge of employing magic to overcome competitors was a standard one in the professional rivalries of the age. Libanius himself incurred it four times, once being formally tried and acquitted, and once banished from the imperial court and the city in which it resided. In old age he found himself apparently on the receiving end of a spell, when he was prostrated by headaches, which ceased when the dried corpse of a chameleon was found in his lecture hall, its head between its legs and one forefoot closing its mouth. He recovered when it was removed, and though now convinced that he had been bewitched, he magnanimously made no attempt to find the culprit.¹³ Libanius also composed a model speech, put into the mouth of an imaginary citizen in an eastern Roman city, which described how magicians used both demons and the spirits of the dead as agents to inflict quarrels, poverty, injury and disease upon living humans. The ghosts were helpless servants, but the demons actively delighted in causing harm.¹⁴ This is a fictional example of rhetoric, given to an imagined character, but there is nothing in Libanius’s other writings to show that he would have disagreed with it, and the speaker seems to be a pagan like himself, showing how such beliefs spanned the different religions. One of Libanius’s pupils later became the leading Christian churchman John Chrysostom, who recalled how as a boy he had almost been caught up in a hunt conducted by soldiers for ceremonial magicians at Antioch. He and a friend had fished a book out of the river, inspired by curiosity, and found to their horror that it was a handbook of magic, flung away by its owner to avoid detection. As it was now in their possession, they were themselves in mortal danger of being accused as magicians, and remained so until they found a safe means of getting rid of it in turn.¹⁵

It seems then that at times and places in the fourth century the laws against magic were enacted with great severity, and that both the fear of bewitchment and the fear of accusation of it, and of other forms of magic, could be powerful in this period. What is less easily deduced is that Christianity had any decisive role in producing this situation. Certainly, it sought to profit from the latter, by increasingly demonizing paganism and associating magic with it, but the fourth-century drive against magicians seems to have been a direct development of earlier, pagan, attitudes, and united the various religious groups. It was a natural projection of the increasingly savage hostility towards magic found in the third-century codes, which may itself have been provoked by the new and sophisticated, text-based ceremonial magic which appeared in the period and may have come from Egypt. Neither Ammianus nor Libanius gave Christians any credit for orchestrating the fourth-century persecutions in the name of their faith; rather, Ammianus blamed insecure new imperial dynasties, led by ruthless new men, who appointed other parvenus to conduct investigations designed to root out treason and criminality in their administrative districts. A succession of historians has studied the trials concerned, and, while they have disagreed over the extent to which these were an expression of hostility between different social groups, they have tended to play down the religious factor, seeing Christianity as abetting rather than causing the persecution.¹⁶ The latter slackened during the fifth century, although the grip of Christianity was by then stronger and the laws against magic had been further augmented and codified. No convincing explanation has been proposed for this, and it may be simply that the imperial authorities were too preoccupied with invasion, civil war and heresy, as the western half of the empire collapsed, to give much attention to magicians.¹⁷

Recent studies have examined different ways in which the new religion adapted to, and exploited, contemporary attitudes to magic. One has traced the manner in which Christian leaders in the late Roman Empire such as Augustine, John Chrysostom and Basil of Caesarea (another pupil of Libanius) used old Roman literary tropes of women casting spells to deceive or ensnare men in order to damn magic in general as pagan: another example of the way in which pre-Christian ideas and images could be employed by the new religion for its own purposes.¹⁸ Another has pointed to the ways in which tales recorded by monks between the fourth and seventh centuries reflected and propagated hostility to magicians. The latter are portrayed as working for hire to cause harm to the rivals or enemies of their paymasters, and as being thwarted by good Christians, after which some of them are burnt and others beheaded.¹⁹ Other historians have examined the ways in which Christians used magic itself, mostly relying on Egyptian evidence. Some carried it on with rites, in the manner of the pagan magical papyri, though often in simplified fashion, combining Christian expressions with esoteric figures and names.²⁰ Others attempted to stay within the ground rules laid down by Origen, offering spells to help clients obtain their desires which depended on scriptural quotations, appeals to the true God and his angels, versions of Christian liturgy, and consecrated oil or water. The authors seem often or mostly to have been monks, who thereby fulfilled much the same role as the ancient Egyptian lector priests.²¹

Witchcraft and Magic in the Early and High Middle Ages

There long existed a scholarly belief that the first thousand years of Christian supremacy in Europe witnessed very little witch-hunting. Back in the 1920s a pioneering historian of European magic, the American Lynn Thorndike, commented of the period up to 1300 that ‘of the later witchcraft delusion . . . we have found relatively few harbingers’.²² At the opening of the recent surge of research into the early modern trials, in the late 1960s, the leading British scholar Hugh Trevor-Roper stated firmly that ‘in the Dark Age there was at least no witch-craze’, and that the early modern belief in witches was ‘a new and explosive force’.²³ In the mid-1970s, in a pair of works which laid down much of the agenda for the subsequent quest for the meaning of the early modern trials, his compatriot Norman Cohn and the American Richard Kieckhefer both agreed. The former wrote that ‘there is little positive evidence of maleficium [i.e. witchcraft] trials before 1300’, while the latter concurred that before 1300 ‘the incidence of witchcraft was so rare that it is impossible to detect patterns of accusation’. Certainly, even in the early fourteenth century ‘the rate of prosecution was low indeed’, and diminished still further in the middle decades.²⁴ In 2004, however, Wolfgang Behringer mounted a challenge to this orthodoxy, arguing that early medieval law still prescribed the death penalty for witchcraft, and that the lack of legal records for the time could conceal many trials. He pointed out that chronicles from the period referred to executions of suspected witches across Europe, in some places more frequently than in the early modern period.²⁵ His argument is not wholly polarized against the preceding belief, as Cohn had accepted that while trials were rare, there were some dramatic cases of the lynching of suspects by mobs, which Behringer counted in his record of persecution. Moreover, Behringer himself conceded that there did seem to be a relative lull in action against witchcraft in Western Europe between 1100 and 1300, which he ascribed to an improved climate, which generated greater security. None the less, his challenge to the twentieth-century portrait of relative medieval tolerance has reopened the question of how much witch-hunting actually went on in early and high medieval Europe; and that must now be considered.

Before confronting it directly, it must be acknowledged that the official attitude of early medieval Christians to magic, as defined by orthodox churchmen, was generally uncompromisingly hostile. Following the argument developed by the time of Augustine, it regarded all attempts to wield spiritual power to achieve material ends as demonic unless deployed by its own accredited representatives, and using only prayer, Scripture or its liturgy as instruments. It had, moreover, greatly enlarged the category of demons by consigning to it all the deities of paganism, and its definition of magic, and thus of forbidden uses of ritual power, including most if not all forms of divination and of traditional charms and spells used to heal and protect. It may be pointed out once again that this was in many ways a development of pagan attitudes. Roman emperors had striven to control or banish forms of divination that were not associated with traditional religion, and (as said) increasingly ferocious laws were passed against magicians in general. None the less, it was a development, and an enlargement, summed up in the fact that all forms of magic became officially described as maleficium, a term reserved before for deeds which caused actual harm.²⁶ By an opposite process of linguistic cross-fertilization, the Roman word sortiligium or sortilegium, meaning divination by the casting of lots, transformed in the course of the early Middle Ages into one frequently used for all forms of magic, and in particular for the least reputable kinds, involving the invocation of spirits (and therefore, to the orthodox, of demons). In step with this, the Roman sortiarius, the term for somebody who told fortunes with lots, turned into the Old French sorcerie and through it into the English ‘sorcery’, which had the same broad usage as the medieval Latin sortiligium.

On the ground, things could be different, as the very complaints and invectives of medieval churchmen testify to the fact that magicians continued to flourish even at royal courts, while commoners still resorted in large numbers to lower-grade counterparts for their ills, worries and desires. Between the ninth and twelfth centuries Christian thought itself, at least as articulated by certain writers, adapted to legitimize some forms of divination and healing charms anew. None the less, for the most part the new religion condemned magic in general more fiercely and uncompromisingly than its predecessors had done, and this persisted through the whole medieval period.²⁷ Moreover, Wolfgang Behringer is correct that medieval law codes, starting with those of the Germanic kingdoms which supplanted the western Roman Empire, continued to prescribe penalties for the deliberate working of harmful magic. If the harm done was serious, such as murder, then the penalties were as severe as those specified for doing equivalent damage by physical means; which is logical in societies, such as those in medieval Europe, which believed in the literal potency of spells and curses.²⁸

The results of these laws can be divided into two categories. One consisted of accusations of the use of magic to harm or constrain as a political weapon, of the sort that had been found among the Hittites and in imperial Rome. As such, it was deployed to promote feuds within families, allocate blame for the sudden death or mysterious illness of a ruler, assert the authority of one, or remove a minister. It remained a widespread and persistent, if occasional, feature of the Middle Ages, occurring in the Frankish kingdom in the sixth century and the Frankish empire in the ninth (thrice), at the French royal court and in that of the count of Maine in the tenth, in Aquitaine and Aragon in the eleventh, and in Flanders and Byzantium in the twelfth.²⁹ Fewer than a dozen cases, however, spread over eight hundred years and most of Europe, do not represent a significant element in medieval political life and state-building; and such charges were spectacular enough to be recorded by chroniclers.

The second category of actions against suspected evil magicians consisted of local trials and mob actions, and here again there is a succession of episodes, some long well known.³⁰ Women were put to death for practising magic at Cologne in 1075, Ghent in 1175, in France in 1190 and 1282 and in Austria in 1296. Most of these cases involved one or two victims, but there is a chronicle reference to the burning of thirty women in a single day in the south-east Austrian province of Styria during 1115, for an unrecorded offence that, given the penalty, was probably witchcraft. A priest was burned as a magician in a Westphalian town in about 1200. These all seem to have been legal executions, but in addition there were lynchings, especially a notorious case at Freising in Bavaria in 1090 when three women were burned to death by their peasant neighbours who blamed them for poisoning people and destroying crops by uncanny means. Bishop Agobard of Lyons published a now famous sermon concerning the murder by mobs in the 810s of people suspected of using magic to cause storms that annihilated crops, and epidemics among humans. In Denmark in 1080 women were also blamed for bad weather and disease. Medieval Russia seems to have been particularly prone to such responses, and a string of reports between 1000 and 1300 testify to the killing of old people, with or without legal sanction, by crowds that blamed them for engineering crop failures and so causing famines: this was a reaction which had died out by the early modern period when witch trials were at their height elsewhere in Europe. Around 1080 King Wratislaw II of Bohemia allegedly supported his brother, the bishop of Prague, in punishing individuals accused of using magic to cause madness and storms, and purloin the milk and grain of other farmers. According to this account, he beheaded or burned the male suspects and drowned the women, to a total above a hundred.³¹

Wolfgang Behringer seems right again, therefore, that in some countries murders or executions of people for destructive magic were more common in the allegedly tolerant early and high Middle Ages than during the main, succeeding, period of trials. None the less, the pattern revealed by the surviving records is one of occasional executions or murders of isolated individuals or very small groups in Western Europe, and sporadic savage outbreaks of worse persecution, especially in eastern parts of the Continent, triggered by unusual calamities of hunger or disease. The recorded cases must undoubtedly represent only a proportion of an unknown number that has been lost to history, but the records of them suggest that these were events dramatic and rare enough to be worth chronicling. What seems to be especially significant in those records is the part played by churchmen. As said, they regularly and vehemently condemned all or most kinds of magic as demonically inspired and assisted, and none seems to have argued against a belief in witchcraft: indeed, the most influential theologian of the central Middle Ages, Thomas Aquinas, ruled firmly that the Christian faith proclaimed that to dispute the existence and effectiveness of harmful magic was to deny the reality of demons.³² On the other hand, they did not develop any theology that demanded and encouraged the hunting of witches either, and, in practice, seem to have acted more frequently to encourage than to discourage it before 1300. It is true that the bishop of Beauvais in northern France led the citizens in trying and executing a woman for practising magic in 1190.³³ Against this it must be recorded that we know that women were blamed for storms and epidemics in Denmark in 1080 because the Pope himself, Gregory VII, wrote to the king to stop it, as a barbaric custom which prevented the realization that such disasters were divine punishments.³⁴ When Bishop Agobard preached against the murder of presumed storm-raisers and disease-bringers, it was to denounce the practice vehemently on the same grounds. Agobard claimed to have intervened to save the lives of some of those accused, while noting that the persecution had to some extent been provoked by men who tried to work a protection racket on farmers, demanding money from them to spare their crops from being magically attacked.³⁵ One of the Russian persecutions of people for causing famine, in the diocese of Vladimir in the 1270s, was recorded because the bishop condemned it with the same theological argument as Pope Gregory.³⁶ The three women murdered at Freising in 1090 only suffered because the local bishop had died and a successor had not been appointed, creating a gap in formal authority, and monks from a nearby religious house interred the burned remains in their own cemetery, as those of martyrs.³⁷

Such clerical attitudes would perhaps have done much to discourage precisely that popular tendency to blame magic for natural disasters, which seem to have had the greatest potential to generate local witch-hunts in the period. It appears to have been of a piece with them that when, in the thirteenth century, the Catholic Church developed a formidable inquisitorial machinery to detect and annihilate heresy – the holding of false religious opinions – Pope Alexander IV ruled that magic in itself should not be the concern of inquisitors.³⁸ Dominican friars were the most active staff of the new inquisitions, but in 1279 some of them stopped the burning of a woman as a witch by peasants in Alsace.³⁹ Churchmen between 500 and 1300 were also generally consistent in condemning as illusions and superstitions some widely held popular beliefs which would, had they taken them literally instead, have encouraged witch-hunting. One of these was the belief in the night rides undertaken by followers of Diana or Herodias, or the other names attributed to the roving superhuman lady or ladies, declared illusory from the canon Episcopi onwards. Another belief was that in night-roaming cannibal women who attacked sleeping people, children or adults, and consumed their organs. It was seen in the second chapter of this book that this idea, embodied in the earliest surviving Germanic law code, was outlawed in subsequent codes as the effect of Christianity (and perhaps of educated Roman opinion, which had questioned the reality of the strix demon from pagan times) was felt. Early medieval sermons and penitentials continued to condemn belief in such figures as a fiction.⁴⁰ Thus it can be suggested that early and high medieval churchmen both believed in the existence of magicians – and indeed this was beyond doubt as there were always plenty of people offering magical services, and also probably some who did attempt to use magic to harm enemies – and the need to stop them; and yet operated in many ways to reduce the likelihood of frequent and large-scale witch-hunting.

What at first sight appears to be a notable exception to this rule turns out on closer inspection to prove it: the case of Archbishop Hincmar of Rheims, who in 860 came to the defence of Theutberga, the wife of the emperor Lothar II, whom her husband wished to divorce. He accused the supporters of the annulment, and especially Lothar’s mistress Waldburga, of using magic to further their ends. He made clear that he believed their powers to be real, and gained in alliance with demons. In accordance with the custom of the time, he lumped together practices such as divination with those designed to afflict others, in the same general category of evil and forbidden craft, to be fought with the rites of the Church. Moreover he believed that the workers of magic were usually female, and motivated by a desire for power over men. At no point, however, did he call for their trial and execution, so confident was he that the effects of Christian ritual, and especially of consecrated salt and oil, were sufficient to undo their work without need of any further measures.⁴¹ In addition, Norman Cohn drew attention to the importance of the medieval legal system itself in damping down accusations, by confronting those who accused others of crimes with a substantial penalty if they lost their cases. As the use of magic was in its very nature hard to demonstrate, this could present a formidable obstacle.⁴² Cohn’s argument is almost certainly correct in as far as it concerns private charges brought between individuals (although, as shall be seen, it was not impossible to provide evidence which could convince a court even under these conditions). What is clear, however, is that such barriers to prosecution and conviction could be flimsy indeed if large numbers of people in a community rounded on presumed workers of magic and believed the charges against them, or if a ruler gave such an accusation full credence. This was true under the Romans, who developed the legal system concerned, and remained so as long as that system endured; otherwise the cases listed above would not have occurred. It may be suggested, therefore, that a will to hunt, at least regularly and intensely, was what was missing, and that allowed a judicial process unfavourable to accusations of witchcraft to persist. Wolfgang Behringer’s suggestion that the climatic upturn of the high Middle Ages left people more prosperous and secure, and so less likely to fear witchcraft, may have some bearing on the relative absence of trials and murders at that time. The extent of decline was, however, slight, given that such events had always been few; and the idea is somewhat difficult to prove. A warmer climate is not necessarily less stormy or unhealthy, and the witch-hunts in Russia were provoked by famines caused by droughts, not floods, and continued through the climatic optimum. It may be that, once again, ideological factors were more significant than functional factors in producing a low and intermittent rate of persecution of presumed witches. After all, exactly the same period, of the twelfth and thirteenth centuries, saw what has been called ‘the formation of a persecuting society’ in Europe, as its peoples turned upon Jews and homosexuals with a new hostility and introduced ever more rigorous measures and structures to deal with a new problem of widespread and mass Christian heresy.⁴³ The witch figure, however, did not get caught up in these developments. These seem to be the conclusions that can be drawn from a Continent-wide survey, but a closer and more detailed survey of a single country may contribute more to knowledge of the matter; and here England provides an obvious choice, not merely because of convenience to the present author but because of the range and quantity of its surviving records. Here as elsewhere, early medieval churchmen denounced magic in all its forms, and prescribed penances for the practice of it, though it is now difficult to distinguish destructive forms of it because of the imprecision of the terminology used. As said before, the Latin maleficium could now be applied to neutral or even beneficial objectives if magical operations were used to secure them, and Anglo-Saxon expressions (to be discussed below) were apparently used as broadly in clerical texts.⁴⁴ Things are only a little clearer when turning to secular law codes written in the native language. The latter had more than thirty terms for magical practices and practitioners, but a few recur with especial frequency, and were criminalized most often. Their meaning can only be recovered vaguely, and by association with other words in Anglo-Saxon that include their components. Gaeldorcraeft seems to have had connotations of song or incantation; libcraeft of potions; and scincraeft of delusion and phantasm. That leaves wiccecraeft, the ancestor of ‘witchcraft’, practised by a female wicce or male wicca, from which of course comes ‘witch’: as the ‘cc’ was pronounced as ‘ch’, the derivation is even closer than the spelling suggests. This is more difficult to match with other expressions, and so we rely on context to reconstruct a meaning.⁴⁵ Such a meaning does seem to emerge from the criminal law codes. The earliest, of Alfred the Great at the end of the ninth century, is also the least helpful, because as a self-consciously devout Christian king he paraphrased the Hebrew injunction in the Book of Exodus. Thus, gealdorcraeftigan, scinlaecan and wiccan, and those who resorted to them, were not to be allowed to live; but Alfred provided no further definitions of these (and the original biblical reference is itself, as discussed earlier, obscure). His law was augmented by one issued by his grandson Athelstan in the late 920s, which made death the penalty for killing by wiccecraeftum, lyblacum and morðdaedum: in other words by the three dishonourable means which were alternatives to homicide committed in a fair fight: magic, poison and murder by stealth. This set of provisions was further amplified by a series of law codes issued by successive regimes between 1000 and 1022 which all seem to have been drafted by the same reforming churchman, Archbishop Wulfstan II of York. They prescribed banishment, or death on refusal to depart, for wiccan and wigleras (the latter being another term for a magician). The significant thing here is the list of other offenders placed in the same clause: perjurers, murderers by stealth and either prostitutes or flagrant serial adultresses. All these could be regarded, once more, as committers of offences against the person, and so wiccan and wigleras were probably, in this context, workers of destructive magic. The clause was repeated, with some amendments such as the inclusion of scincraeft or libcraeft, in the other codes drafted by Wulfstan.⁴⁶

There does not appear to have been any stereotypical witch figure in Anglo-Saxon England: people seem just to have been expected to yield to the temptation to use magic against their fellow humans at particular moments and for particular reasons. The terminology applied to magicians also suggests that there were individuals who had a reputation for expertise in particular kinds of magical technique, who might provide their services for malicious as well as beneficial ends; and the terms show that these were expected to be of both sexes. To churchmen, of course, all magical practitioners were probably colluding, consciously or not, with demons, and persisting in ways which came out of a pagan past in which these devils were openly worshipped; but the concern of clerics was mainly with condemning acts of healing and divination, which still attracted much belief and support from the populace, and they paid little attention to destructive magic.⁴⁷ There are traces of belief in supernatural female figures who could protect or harm humans, but especially the latter. Again, different terms were used for these, most of them rarely encountered and now difficult to understand, but one, haegtis or haegtesse, occurs more commonly in texts, and was to engender the word hag, for a malevolent old human woman.⁴⁸ Early English glossaries equate it with the Latin word striga, for the murderous night-roaming female demon.⁴⁹ One now very famous healing charm, ‘against a sudden stitch’, brings such figures momentarily sharply into focus, as ‘mighty wives’, who ride across the land yelling and sending darts which cause stabbing pain in the humans at whom they are aimed. They are equated with both elves and pagan deities, so are clearly not supposed to be mortal. The charm aims a magical spear back against them in turn.⁵⁰

What is missing in the early English texts is the cannibal woman, who preys upon people in the night, found in the first Germanic law codes. It could be that Christianity had stamped out the belief in her before the sources mentioning hags began to be written; but a number of pagan elements, including hags themselves, got through into those. The figures in the charm sound a lot more like the night-riding women of medieval Scandinavian literature – though there is no sign in the charm that the pain-inflicting females there are thought to be active specially by night – and indeed this northern literature, produced by peoples who neighboured the original Anglo-Saxons, likewise has no stereotypical witch figure in it. It could be that the concept of the cannibal night-witch was confined to certain Germanic tribes. In addition, there is a trace in one Anglo-Saxon text of a belief in people who have intimate dealings with demons, to the detriment of their fellows: it prescribes a protective salve against the ‘elf-kin’, ‘night-walkers’ and ‘people with whom the devil has sex’.⁵¹ It may be suggested, therefore, that most of the components of the early modern idea of witchcraft were already present in Anglo-Saxon England, but were as yet far from being assembled into that later construct. Of actual signs of enforcement of the laws against witchcraft, there is only one, now celebrated, case. It is that of a tenth-century widow living in Northamptonshire, who drove an iron pin into an image of a man whom she and her son disliked. When suspicion fell on her, her room was searched and the image found, and this proof was sufficient to get her executed, by drowning in a river: her son fled and was outlawed and their land confiscated.⁵²

This record is only preserved, however, because the land concerned swiftly became part of another transaction, in which its new ownership needed to be explained. It is impossible to tell simply from the lack of other evidence whether such cases were as rare as the written sources would suggest, or were routine but, for lack of the keeping of legal records, only likely to emerge into the light of history by accident, as happened in this instance. It may, however, be possible to suggest an answer to this problem by approaching it from other angles. Anglo-Saxon royal families were divided by often vicious internal rivalries, and yet there is no contemporary or near-contemporary mention of magic being used as a charge in these, as it was occasionally in similar dynastic feuds on the Continent. It is also revealing to consult the very large collections of healing and protective charms and herbal medicine that survive from Anglo-Saxon England. Only four such charms were designed, even in part, as remedies against malicious human magic, and non-human beings, such as elves, demons and hags, and depersonalized menaces such as the ‘flying venom’ which caused disease, appear to have provoked more concern.⁵³ It seems that the early English may be counted among the peoples who credited most uncanny misfortune to other sources than their fellow humans; which would have damped down an impulse to hunt witches. Not much altered after the Norman Conquest with respect to attitudes to magic, despite the tremendous political, social and cultural changes it brought in general. Twelfth-century authors did occasionally produce portraits of wicked English female magicians: in total, of a queen who used potions to turn herself into a mare; a woman hired by a stepmother to brew an evil potion to be administered to a rightful heir; another at Berkeley in Gloucestershire, who knew spells and could learn the future by listening to birds; and another hired by the Normans to help them against Saxon rebels, and who delivered a curse (ineffectually) from a platform. All these were, however, literary constructs designed to carry moral messages, set in a past age before that of the writer and involving formulaic activities, so that the image of the woman on the platform, for example, may have been borrowed from Norse accounts of seiđr.⁵⁴ The law itself altered little, William the Conqueror himself forbidding the use of spells to kill people or animals, while a treatise written in the reign of his son Henry confirmed that the penalty for achieving this should be death, but for merely attempting it those convicted should pay compensation.⁵⁵ Legal records exist from the late twelfth century onwards and show a few trials for magic: a woman arrested for it in Essex or Hertfordshire in 1168 and a Norfolk woman tried and acquitted in 1199 or 1209; while in 1280 an abbot of Selby, Yorkshire, was accused of employing a male magician to find the body of his drowned brother. There is no indication in the first two cases that destructive magic was at issue, and in the third it was not, but a thirteenth-century man at King’s Lynn was fined for drawing blood from a woman upon false suspicions; and in later centuries this was a standard means of averting bewitchment. In Northumberland in 1279 a man’s goods were confiscated because he had struck and killed a woman. Her body had been burned, probably indicating that she was believed to have been a witch.⁵⁶ In the late twelfth century Henry II banished all magicians from his court, and by the thirteenth the charge of using magic for personal gain had at last entered the toolkit of high political rivalry: in the first half of the century the main royal minister, and in the second half a high financial officer, the Chamberlain of the Exchequer, were accused of it.⁵⁷ So the fear of witchcraft remained, with laws against it, but does not seem to have been intense or to have manifested frequently in action. The English case seems to prove the European rule for the early to high Middle Ages.

When all this is said, there is a general agreement among historians that official attitudes to magic underwent a significant change in Western Europe during the decades around 1300, because of the impact of the elaborate and ceremonial variety that first appears in late antique Egypt.⁵⁸This was imported during the twelfth and thirteenth centuries through the translation of Greek and Arabic texts, and represented a serious challenge to Christian orthodoxy. For one thing, it represented a new form of magic to medieval Western Europeans, unprecedented in its elaboration and sophistication and disproving the long-cherished expectation that continued condemnation of the traditional, simpler kinds of magical remedy would gradually eradicate resort to them or at least confine it to the poorest and least influential levels of society. On the contrary, the newly arrived texts of complex magic relied heavily on transmission of written figures and formulae, and those subsequently developed from them often demanded knowledge of Christian liturgy and clerical conventions of behaviour. Hence they appealed to the most educated, wealthy and sophisticated social groups, and above all to the churchmen who should have been the guardians of religious orthodoxy. While texts that openly required the invocation of demons would always remain outside the likely bounds of officially acceptable practice, those that claimed to manipulate the natural forces of the universe, and above all the influence of heavenly bodies, were far less easy to condemn out of hand. Even the sub-class which recommended the employment of demons sometimes made a direct and reasoned riposte to orthodox teaching by claiming that the proficient magician could compel and control evil spirits and so force them to work for benevolent ends, so striking a resounding blow for Christianity. For two hundred years, learned authors in Western Europe conducted a debate over how far forms of the new complex magic could be assimilated into orthodoxy and be used for human benefit. By the early fourteenth century, however, the majority of them had swung firmly against such a rapprochement and reinstated the Augustinian orthodoxy that all magic was inherently demonic, whether its practitioners were conscious or not that they were working with demons.

This development accompanied and overlapped with another, which was inspired largely by the appearance of widespread heresy in Western Europe between the eleventh and thirteenth centuries and the increasingly savage, and successful, Catholic counter-attack upon it, using crusade and inquisition as its main weapons: it was between 1224 and 1240 that burning came to be adopted as the standard mode of execution for heretics, as it had long been for magicians. They were routinely portrayed as devil-worshippers as part of that counter-attack, and this strategy encouraged an outbreak of political trials between 1300 and 1320 in which prominent individuals and organizations were accused of worshipping Satan in secret, and often ruined as a consequence. King Philip the Fair of France became the most ardent practitioner of this technique, using it to attack a bishop who was one of his own councillors, a pope and then the crusading order of the Knights Templar, and it was continued under his successor Louis X. In England, the bishop of Lichfield was accused in 1303, and suspicion of magic seems to have quickened at a local level, as a woman was banished from Exeter in 1302 for entertaining notorious magicians from South Devon, and in 1311 the bishop of London ordered measures to be taken to curb the growth of fortune-tellers.⁵⁹ Both developments, the condemnation of magic and the escalation of political trials for devil-worship, were accompanied by a growing fear of the power of Satan in the world; which may itself have been generated, and was certainly reinforced, by the twin new threats posed by large-scale heresy and ceremonial magic.⁶⁰

In this manner the scene was set for a direct and comprehensive attack on ceremonial magic, as demonic, launched by Pope John XXII between 1318 and 1326. He was already inclined to use the charge of malicious magic against personal opponents, having had a bishop burned as a result of it in 1317 and going on to deploy it again thereafter. In 1318 he appointed a commission to root out ceremonial magic from his own court at Avignon, and in the 1320s four other trials of alleged magicians were held in different parts of France, some directly encouraged by the pope: churchmen were accused in all of them, although sometimes assisted by lay practitioners. In 1326 John decreed that ceremonial magic had grown to the proportions of a plague, and excommunicated all concerned in it.⁶¹ Magic was thus at last identified directly with heresy. His actions seem to have had knock-on effects, as the use of magic as a political charge returned to nearby royal courts, once in that of England and twice in that of France between 1327 and 1331; while a woman was burned in the south-west German province of Swabia in 1322 for using a consecrated communion wafer in a magical rite.⁶² Pope John’s influence reached as far as Ireland, where one of his protégés became bishop of Ossory and provoked a sensational and subsequently notorious trial at Kilkenny in 1324–5. Twelve people were accused, of whom the most prominent was Alice, Lady Kyteler. The charges arose out of a feud within a prominent local family, and became the subject of a power struggle between different factions in Church and state among the English settlers in Ireland. In the short term the bishop won this, and the accused were convicted of being ‘heretic sorcerers’⁶³ who abandoned Christianity to worship demons, and gained from them the ability to obtain their own desires, which included the injury and murder of selected human victims. Some of them, including Lady Alice, escaped by fleeing, and others were absolved upon doing penance, but one woman, Petronilla of Meath, was tortured into confession and then burned to death, the first person in Ireland to suffer this fate for heresy.⁶⁴

In 1331 the English royal council ordered a hunt for magicians in London, and three goldsmiths were caught in the act of a magical ceremony in the suburb of Southwark: one was a semi-professional, hired by the others, and he and his main accomplice were remanded in custody while their bishop was consulted about whether their actions were heretical.⁶⁵ The bishop concerned was that of Winchester, whose jurisdiction extended to Southwark, and the case seems to have triggered a wider crackdown on magic in his diocese, where two more trials were held, over the next six years, of villagers who had sought or provided magical aid. The punishments were confined to whipping, however, while the Southwark magus was exiled.⁶⁶ The next pope, Benedict XII, had himself been a notable inquisitor, and an avid hunter of heresy and magical practices, and in 1336–7 he took a personal interest in legal cases involving magic in various parts of Italy and France.⁶⁷ At about that time an Italian professor of theology, Bartolus of Sassoferrato, redefined the ancient terms for child-killing nocturnal demonesses, striga and lamia, to mean ‘a woman who renounces Christianity, and deserves death’.⁶⁸

There was, however, no sustained momentum behind this sequence of persecution. Benedict seems to have been as interested in ensuring that justice was done in the cases in which he involved himself as in pursuing the accused, and in one instance he directed fresh investigation of a charge of using image-magic to kill John XXII, which the latter had believed but which Benedict thought could be fraudulent.⁶⁹ The pope who followed him in 1342, Clement VI, was apparently not concerned with the issue, while the French and English royal families entered a period of internal unity and stability. No local tradition seems to have been established of a popular fear of magic-workers or a crusade against them by inquisitors or secular magistrates, and so prosecutions seem to have died away at all levels of society. This makes their revival from the mid-1370s all the more remarkable. It was not led by the papacy, which in 1378 ruptured into the Great Schism, forty years in which rival popes strove for supremacy with the different Catholic states supporting one or the other. One of the last pontiffs before the division occurred, Gregory XI, was asked in 1374 by the chief inquisitor for France for renewed powers to repress ceremonial magic, which was allegedly rife and attracting priests: he thought that no such authority remained to him from the earlier period. Gregory gave them, but only for two years.⁷⁰

There is more purchase for the idea that renewed political insecurity readmitted the charge of magic to high-level dirty politics, which put it back into the limelight. In 1377 England found itself with a senile king besotted with a mistress, and a child heir. The mistress was promptly accused of using spells to gain the old king’s love. When that child heir succeeded, one of his ministers was accused of demonic magic after that man’s fall and execution. The young king concerned, Richard II, was subsequently deposed, inaugurating a long period of turbulence in English dynastic politics which culminated in the Wars of the Roses; and five out of the six reigns between 1411 and 1509 were marked by at least one accusation against somebody, usually a member of the royal family, of using magic to try to kill the current monarch.⁷¹ In the 1390s the reigning king of France went mad, and this was blamed on witchcraft, especially as the resulting power vacuum engendered a particularly vicious and prolonged struggle between other members of his family in which the same charge played a prominent part; as it continued to do when that struggle led to the collapse of France into civil war during the following decades. By 1398 it had already provoked the University of Paris into reaffirming the doctrine that ceremonial magic, as assisted by devils, constituted heresy, and further discussions and condemnations of it by intellectuals associated with the warring parties followed.⁷² The duke of neighbouring Savoy duly claimed to uncover a magical murder plot against himself in 1417.⁷³

Most of this renewed royal, aristocratic and scholarly interest seems, however, to have followed and become meshed with a new hostility to magic at a local, and especially an urban, level. In 1376 the inquisitor for Aragon in Spain, a Dominican friar called Nicholas Eymeric, issued what was to become an immensely influential handbook for the definition and detection of heresy.⁷⁴ This made the first unambiguous declaration that ceremonial magicians were to be regarded, hunted down and punished as heretics. Even Eymeric ruled that some of the simpler practices of service magicians which did not require the conjuration of spirits, such as reading palms and drawing lots, were not to be the concern of inquisitors, but in practice the new drive against magicians, of which his manual seems as much a symptom as a cause, sometimes seems to have elided the two. In 1390 the Parlement of Paris declared sorcery to be an offence within its own jurisdiction, and proceeded subsequently to try two women from the city for attempting to work love spells, a third who had offered a range of magic to customers and fourth who had tried to use it against her abusive husband, and burned them all for devil-worship.⁷⁵ The same thing happened in the same year at Milan to two women who had sold magical services to clients, services which they claimed to have learned from the superhuman ‘lady’ whom they followed by night.⁷⁶ A confession of diabolism was likewise wrested from a woman at Geneva in 1401, who had claimed to consult a spirit in order to help clients find stolen goods and protect livestock.⁷⁷ There were no cases of magic in the secular courts of Florence between 1343 and 1375, but three convictions and two executions between 1375 and 1412: all seem to have been of people who had practised it for their own ends or offered to perform it for a fee.⁷⁸ The same courts at Lucca tried nobody for magic between 1346 and 1388, yet convicted three between 1388 and 1415, two of them foreigners offering services for hire.⁷⁹ The element of diabolism appeared occasionally in cases in both cities. In London, people who offered magic for hire were punished in the 1390s and 1400s, and the bishop of Lincoln received a royal order to do the same to all in his diocese in 1406.⁸⁰

The reasons for this upsurge in accusation and prosecution, across Western Europe and at different levels of society, may not be possible to discern confidently in the present state of knowledge. Michael Bailey has noted the number of treatises published between 1405 and 1425, by French and German scholars, which applied demonological theories to simple and mundane spells and charms; and related these to a broader move among churchmen to a practical and pastoral, rather than a cosmological, theology.⁸¹ The groundswell of such a movement may have helped create the conditions for the renewed persecution of magicians, though the texts concerned are all too late to have played a part in starting it. It is easy to believe that most of the people accused of practising ceremonial magic actually did so, because many examples of it have survived, which contain rites and spells, both to help the practitioner and to hamper or injure enemies, similar to those cited in the court records. It is also quite credible that some practitioners would actually have invoked demons, as those surviving texts of ritual magic sometimes contain instructions on how to do so; the presumption, of course, being that the magician would be constraining the devils concerned to his or her own will.⁸² When a Greek woman tried at Lucca in 1388 is recorded as summoning infernal spirits in the names of God and the Virgin Mary, to aid in rites to gratify her clients, there is no paradox in the statement: she would have been using these holy names to gain power over the demons concerned.⁸³

What is much more in doubt is that any of those tried in this period actually worshipped Satan or his lesser devils, as some were convicted of doing. For the charge of heresy to stick to magicians, this is what they had to admit. Norman Cohn made a convincing argument that no widespread sect of Satanist magicians existed.⁸⁴ It is considerably harder to determine whether or not individuals, or even small groups like that around Alice Kyteler, forsook Christianity to give allegiance to the Devil or a devil. It would have been against the whole tradition of ceremonial magic, as expressed in its known texts, to do so; but the existing evidence is not adequate to suggest any final answer to the problem. What it does strongly suggest is that some of the magical practitioners who were accused claimed to have relationships with spirits, as helpers or servants, which those interrogating them turned into demons; but how far this explains the charges of devil-worship in general is, again, hard to decide.

There is, on the face of things, no reason why the upsurge in official attacks on magic at the end of the fourteenth century should not have subsided as those in the early part of the century had done. Instead it blended seamlessly into what turned out to be the beginning of the early modern European witch-hunt. In 1409 one of the contending popes in the still persistent schism, Alexander V, sent a decree to the inquisitor general whose territory covered the western Alps, ordering him to proceed against new forms of deviance which practised heresy, usury and magic there. The definition of magic included the elaborate literary kind, divination and peasant superstitions: so if the groups that practised them were thought to be new, there is no real evidence that what they did was regarded as novel.⁸⁵ The document was probably sought by the inquisitor himself, and there is nothing out of the ordinary about it: its treatment of magic fits into the general crackdown of the age and is a papal equivalent to (for example) the order sent to the bishop of Lincoln, so that the pope concerned was belatedly following current trends rather than leading them. What makes it more significant is that the inquisitor in question was Ponce Feugeyron, a Franciscan who less than three decades later was to be involved in some of the earliest witch trials of the early modern kind.

It may be helpful at this point, therefore, to emphasize how the image of the magician which underpinned the fourteenth-century trials differed from that of the satanic witch which underpinned those of the early modern period. There was no sense in the late medieval attack on magic that magicians were part of an organized and widespread new religious sect, which posed a serious menace to Christianity. They were, rather, viewed just as individuals or small individual groups, in particular places at particular times, who yielded to the temptation to gain access to normally superhuman powers for their own ends. The ends concerned, though selfish, were generally just for personal profit rather than dedicated to the commission of evil as an end in itself, and most of those targeted offered their services for sale to others or sought assistance from such experts. The acts with which they were charged were usually heavy in the paraphernalia – special objects, substances and spoken words – on which ceremonial magic generally relied. In most cases the element of apostasy from Christianity was not central to the charges, and because those accused were not expected to belong to a sect, there was no cumulative effect of arrests, as those already under interrogation were not required to name accomplices. As a result of all these features, the overall body count produced by the persecution was low: between 1375 and 1420 the total number of people executed for offences related to magic, across Western Europe, was probably in the scores rather than hundreds. In this period as throughout the previous Middle Ages, there was in practice no significant element of gender among those tried, save that – mirroring educational patterns in society as a whole – men were more likely to be accused of the more text-based and learned kinds of magic, and women of the less. The stereotype of a witch that underlay the early modern trials had not yet appeared by the opening of the fifteenth century.

The Making of the Early Modern Witch

The most important feature of the concept of the satanic witch that appeared at the end of the Middle Ages was that it was new. This was, as shall be seen, fully acknowledged at the time of its appearance. In 1835 Jacob Grimm, as part of his pioneering work into the history of Germanic folklore, came up with a two-stranded explanation for its development. One strand, the more dynamic, consisted of the increasing concern of the medieval Catholic Church to purify the societies which it controlled, by identifying and eliminating heresy. This supplied the basis for the imagination of an organized sect of devil-worshipping witches, but Grimm also suggested that the forms which that imaginative creation took were conditioned by his second strand, popular beliefs inherited ultimately from the pagan ancient world.⁸⁶

It will be the contention here that Grimm’s model was correct, but it must be noted too that historians developed its two components largely separately in the century and a half after he wrote. That of folk beliefs rooted in the ancient world acquired momentum in the rather different form of seeing the people prosecuted as witches as practitioners of a surviving pagan religion, which, as said earlier, finally ran into a dead end in the 1970s. Scholars who were themselves expert in the later Middle Ages and early modern period tended to emphasize the other component, of a medieval Western Church determined to identify and eradicate heresy, as not merely false but as satanic religion. One of the greatest of these was an archivist from Cologne, Joseph Hansen, who at the beginning of the twentieth century edited and published many of the primary texts relating to the medieval and early modern persecution of magicians and witches. His collections have been an invaluable resource for historians ever since, and are prominent in the endnotes to this present chapter. It was also he who pinpointed the apparent place and time at which the stereotype of a satanic sect of witches first appeared: in the western Alps during the early fifteenth century.⁸⁷ The collapse of the theory of a surviving pagan religion, and the beginning of sustained and large-scale research into the witch trials, cleared the way for a new investigation of the origins of that stereotype. In circumstances considered in the last chapter, one was provided in 1975 by Norman Cohn, who essentially restated Grimm’s explanatory model, starting afresh and with much more extensive evidence.⁸⁸ Once more the primacy of importance of the Church’s drive against heresy was stated, but with much deeper roots, going back to ancient Roman stereotypes of antisocial behaviour by groups with aberrant religious beliefs, especially early Christians themselves. He also drew attention to the orthodox reaction to ceremonial magic in creating the particular context for a drive against witchcraft. None the less, Cohn also emphasized the significance of the folkloric element, likewise rooted in antiquity, in contributing important images to the new concept of the witch. His interpretative model was robust and convincing enough to secure general assent. In 2004 one author on the subject, Steven Marrone, could commence a study of his own by declaring that ‘there is no need to rehearse Cohn’s argument here or re-examine his evidence. Both have been so well received as to constitute a fixture of current understanding of the rise of the witch-craze.’⁸⁹

This being so, two other first-rate historians who were also working on the same problem, Richard Kieckhefer and Carlo Ginzburg, had to signal their differences from Cohn in order to draw attention to the value of their own ideas. Both actually endorsed his basic model, of witch-hunting as a spin-off from heretic-hunting but informed by folkloric traditions. All three of them agreed also with Hansen’s identified place and time as the most important origin-point for the new concept of witchcraft. Kieckhefer indeed only disagreed with Cohn over details, tackled in footnotes, but Ginzburg was much more emphatic in drawing attention to his differences from both Cohn, and Kieckhefer.⁹⁰ Those differences embraced both parts of Cohn’s (and Grimm’s) model. With respect to the folkloric elements, Ginzburg related those which had been found relevant before to a much broader and deeper cultural substratum in ancient Europe which he termed shamanistic; and which has been considered at length in Chapter Three of the present book. With respect to the heresy-hunting element, he placed a new emphasis on the importance of specific persecutions of lepers and Jews as secret enemies of society in fourteenth-century France, in preparing the way for a novel stereotype of witchcraft. In 1996 Michael Bailey summed up Ginzburg’s book as ‘one of the most controversial studies of witchcraft’, and its status in this respect has not much altered.⁹¹ The inspiration it gave to authors such as Éva Pócs and Wolfgang Behringer, considered earlier, is not relevant here, because they were really concerned with the manner in which folkloric motifs coloured local witch trials, and not with the origins of the stereotype of the satanic witch itself. Historians who have been concerned with those origins have tended instead to elaborate the part of Cohn’s model that referred to heresy-hunting. Michael Bailey has found the conception of the early modern witch in the conflation by clergymen of elite ceremonial magic, the common tradition of practical spells, and the general fear of malevolent magic, in a single demonic construct. This mixture was then grafted onto standard medieval caricatures of heretical sects. Bailey faulted Ginzburg for overstating the factor of night flight in the creation of the image of demonic witchcraft, such flight being a crucial element in Ginzburg’s argument for the importance of shamanistic traditions in that process of creation.⁹² Bailey’s fellow American, Steven Marrone, has emphasized the impact of ceremonial magic, and the greater agency which orthodox churchmen allowed to demons in response to that impact.⁹³ Wolfgang Behringer and the Swiss historian Kathrin Utz Tremp have restated the importance of attitudes to heresy, showing how trials of heretics in parts of the western Alps turned seamlessly into trials of satanic witches.⁹⁴ The Dutch historical anthropologist Willem de Blécourt has rejected Ginzburg’s model comprehensively, arguing that the shamanistic analogy is completely unhelpful and that Ginzburg had made an inappropriate projection of atypical south-eastern European folk customs onto Western Europe.⁹⁵It has actually been Richard Kieckhefer who has applied Ginzburg’s ideas most closely to the question of origins, agreeing that popular mythologies were important but suggesting that there was no unified imaginative construct of witchcraft in the fifteenth century; instead he has argued for multiple mythologies, in regional varieties, which functioned differently under different circumstances.⁹⁶

There is thus a considerable recent debate over the matter; but there is also a large amount of new material available with which to take that debate forward. One conclusion that can be drawn from that material is that the western Alps alone were not the birthplace of the construct of the satanic witch. Indeed, the earliest clearly dated reference to that construct is in another range of mountains, the Pyrenees. There, in 1424, the leading men of the Aneu Valley, high in the Catalan end of the range, were summoned by the local count and agreed to act against local people who accompanied bruxas by night to do homage to the Devil. They would then steal sleeping children from their homes and murder them, and use poisonous substances to harm adults. Some had already been apprehended and confessed to this crime, and it was decided that they and any convicted of it in future would be burned to death.⁹⁷

Bruxas was a medieval Catalan term for the nocturnal demons known in ancient times from Italy to Mesopotamia who were believed to kill children: the Roman striges. In the course of the fifteenth century, as the belief in gatherings like those in the Aneu Valley spread slowly across northern Spain, it came to be applied to the women who attended them, until it became the standard Spanish term for witches.⁹⁸ A leading expert in the Spanish material ascribes the appearance of the demonic witch stereotype in Catalonia to the activities of Vincent Ferrer and his disciples, Dominican friars who staged preaching campaigns between central France and north-eastern Spain between 1408 and 1422, calling in particular for the punishment of magicians as part of the new Western European crackdown on them. He notes likewise that from the 1420s onwards, secular courts in the Languedoc area of France, bordering Catalonia, began to prosecute individual women for doing homage to the Devil and thus acquiring the power to enter houses through closed doors and poison the inhabitants.⁹⁹

In the same year in which the stereotype of a sect of satanic witches appeared in the Pyrenees, it surfaced at Rome itself, where two women were executed for killing large numbers of children by sucking their blood on the orders of the Devil. They gained entry to the homes of their victims by anointing themselves with ointments and turning themselves into cats.¹⁰⁰ More details of the sect were provided in a trial held by the captain of the city of Todi, to the north, in 1428.¹⁰¹ It was of a celebrated local service magician who sold spells and charms to secure health and love, and break bewitchments. She became caught up in the continuing drive against magic, but what was new about her case was that she was also charged with sucking the blood and life force of children, like a strix, when going abroad at night in the shape of a fly. Moreover, she was accused of riding a demon in the form of a goat (when herself in human form) to join other people of her kind in revelling and worshipping Lucifer, who ordered her to destroy the children. To make it possible for her to fly on the demon’s back, she was supposed to anoint herself with substances such as the blood of babies, and of bats, and the fat of vultures. She was sentenced to burn. Once again a preaching campaign has been associated with this case, this time that of Bernadino of Siena which covered central Italy between 1424 and 1426 and directly encouraged his audiences to report practitioners of magic to the authorities. He had spoken at Todi and co-operated with the reigning pope in launching the hunt at Rome. Bernadino did not believe in the reality of gatherings of witches to which participants flew in order to worship Satan, or that the spells of witches had any power over virtuous Christians, or that they could transform their shape into that of animals. In opposing these ideas, he remained an early medieval churchman. He did, however, think that demons acted for the witches with whom they made pacts, taking the form of animals and sucking the blood of babies to kill them; and in this manner he unleashed ancient and widespread fears which earlier clerics had damped down.¹⁰²

Between Bernadino’s territory of central Italy and that of Vincent Ferrer and his pupils, which extended from eastern Spain to the River Rhone, stretched the western Alps, recognized since Hansen’s time as the birthplace of the early modern witch-hunt. The three areas were all connected by the networks of preaching friars: Bernadino, for example, cited a fellow Franciscan who had told him of a group of child-murdering heretics in Piedmont, at the north-west end of Italy, who used the bodies for a potion which conferred invisibility.¹⁰³ The Alpine one was, however, to be the most influential in propagating the new image of witchcraft. Understanding what happened there has become much easier in recent years because of a remarkable cluster of scholars centred on the Swiss university of Lausanne, which lies at the centre of the main region for early witch trials. This group has edited and published the surviving records of those trials, with the literary texts associated with them.¹⁰⁴ The import of their work suggests that the appearance of the new image of witchcraft in the Alps can first be securely dated in the region to 1428, the year of the Todi case and four years after the hunts in the Aneu Valley and at Rome. This was when a vicious series of prosecutions began in the Valais region, in the heart of the western Alps to the east of Lake Leman. It lay at an intersection of linguistic, cultural and political boundaries, where the French-, German- and Romansh-speaking areas of Switzerland met, and with them a complex of territories ruled by the local bishop (of Sion or Valais), the duke of Savoy, and other petty states. The trials started in two French-speaking valleys, Anniviers and Hérens, as high in the Alps as Aneu is in the Pyrenees, but they spread across most of the region. They were recorded, about a decade later, by Hans Fründ, a chronicler at the city of Luzern to the north, who was clearly well informed about events in Valais.¹⁰⁵ What Fründ recorded was the finding of a conspiracy of ‘sorcerers’¹⁰⁶ to kill their fellow humans at the behest of the Devil, whom they worshipped after he had transported them to nocturnal group meetings on chairs into which a flying ointment had been rubbed. Satan, who manifested in animal form, turned some into wolves, to kill sheep, gave others herbs which made them invisible, and changed the appearances of yet others into those of innocent people. Aided by him, they murdered, paralyzed and blinded their neighbours, and produced miscarriages and impotence among them, as well as destroying their crops, stealing milk from their cows, and rendering their wagons and carts useless. In particular, they killed their children by making them sicken at night, so that they could then dig up and eat the bodies. This conspiracy was said to have been growing so fast that its adherents believed they would have taken over the area, and destroyed Christianity, after another year.

Fründ made clear in this case what may be suspected in those of the Aneu Valley, Rome and Todi: that the confessions were extracted by torture, sometimes applied so brutally that people died under it. He estimated that the hunt resulted in the burning of more than two hundred individuals, both female and male, in one and a half years; which is a large body count even by the standards of the early modern witch trials at their height. It was probably the largest for at least a millennium of people put to death for working magic. Local legal records show that it was conducted by petty lords who were driven by a sudden popular fear of witchcraft, and that the trials began in 1427 and lasted until 1436. These records mention all the details recorded by Fründ except the flight to meetings, and make his estimate of the number of executions credible.¹⁰⁷ Once again, a preaching campaign to raise awareness of the threat from magic seems likely to have provided the context for the trials. These mountain valleys may have been remote from the main centres of contemporary population, but were not sleepy places ignored by political and religious authorities: on the contrary, they were on the front line of religious evangelism and state-building.¹⁰⁸ Two aggressive states in particular, the duchy of Savoy and the city of Berne, were seeking to extend their power in the western Alps, while vassals like the landowners in the territory of the bishop of Sion were trying to assert independence of their overlords. In the process the economy was being shifted from self-sufficiency to production for the market, with proportionate social tension. Moreover, mountains had become refuges for members of heretical Christian sects driven from or wiped out in more accessible areas, and so by the late Middle Ages were especial targets for the friars who acted as evangelists and inquisitors, supported by religious and secular authorities. In the Pyrenees, these heretics were above all the Cathars, and in the western Alps another austere and idealistic branch of unorthodox Christianity, the Waldensians. The latter were subjected to especially intense persecution in the western Alpine region in the late fourteenth and early fifteenth centuries, and the trials of them merged into those of the new kind of witchcraft, as both were accused of worshipping the Devil in groups with similar rites; indeed, in parts of the region, the same word, Vaudois, was used of both heresies.¹⁰⁹ The territory in which Ponce Feugeyron acted as inquisitor general, and in which the pope in 1409 had ordered him to root out heresies, including specifically any associated with magic, bordered directly on the Valais. His powers to do this were renewed in 1418 by Martin V, the pope whose election ended the Great Schism and who co-operated later with Bernadino of Siena in the latter’s witch-hunt at Rome.¹¹⁰

The record therefore shows that a belief in a conspiracy of devil-worshipping magicians, to harm other people, and especially to kill babies and children, appeared in the mid- to late 1420s at different points widely dispersed across a broad area, stretching in an arc from north-eastern Spain to central Italy. The single factor which can link them all is the preaching of friars who were co-operating in a campaign against popular heresy and unusually conscious of the danger posed by magic, as part of the resurgence in the prosecution of its practitioners which had commenced in Western Christendom in the 1370s. This seems to have ignited responses among the populace, amounting at times to panics, in particular places where the circumstances were propitious, perhaps because of unusual infant mortality and other misfortunes, and certainly where justice was in the hands of local secular lords and captains who were easily carried away by public feeling, in a period of political and economic instability. There was a clear connection between these responses and folk beliefs derived from ancient origins, but those did not so obviously derive from ‘shamanistic’ motifs of spirit-flight so much as the figure of the child-murdering demoness, the Roman strix and the Germanic nocturnal cannibal woman. The main part of the new construct, of a group of people who gathered secretly by night to worship the Devil, who appeared to them in animal form, was absolutely standard as an orthodox accusation against heretics in the high and later Middle Ages.¹¹¹ Richard Kieckhefer has pointed, correctly, to the differences of detail between the Swiss and Italian cases as mirroring distinctive local folkloric traditions, and argued from them that there was no single imaginative model of satanic witchcraft in the fifteenth century; only multiple regional mythologies.¹¹² It is proposed here instead that there actually was such a single construct involved, right at the beginning, and that it took on local forms as it was propagated. The obvious creators and propagators of it were the preaching friars of the mendicant orders, Dominican and Franciscan. Moreover, they were not just any members of those orders, but leaders of a particular movement within them, the Observant, which believed in purging Christendom of all laxity and ungodliness as part of the period of reform, which succeeded the Great Schism of rival popes that had riven the Western Church in the decades around 1400. A study of pastoral literature published in and around Siena in the late fourteenth and early fifteenth centuries provides an important microcosm of this movement at work.¹¹³ It reveals that most of it, following earlier medieval tradition, treated popular magic as the product of ignorance and credulity and not as heresy, saw harmful magic as antisocial rather than demonic, and condemned belief in the strix. Bernadino took the opposite view on the first two counts and identified the strix as a human woman empowered to fly by the Devil. In the process, he and his fellow reformist preachers gave a new power and terror to the image of the strix and launched a new kind of witch-hunt. The bricolage of elements that went into the creation of that new model is clearly displayed in the second of the famous texts which have recently been edited by the Lausanne cluster, the Formicarius (Ant Hill) of Johannes Nider, a prominent Dominican, written in the 1430s.¹¹⁴ Nider relied on others for his knowledge, and combined into his model of satanic witchcraft phenomena reported to him by three different people. One was a former ceremonial magician, who had hired out his services, and another a Dominican inquisitor at Autun in central France, who described the local magicians, many apparently serving clients, whom he had prosecuted. The latter man also reported the exposure of a sect of devil-worshipping sorcerers in an area near Lausanne and the Duchy of Savoy, which was presumably the Valais. Most interesting to historians, however, has been Nider’s third source, a ‘Judge Peter’ from Berne, who governed the Simmen Valley on behalf of his city, which had just annexed it. This valley ran into the mountains south of Berne, which divided its region from the Valais. ‘Peter’ became concerned about magic because of a panic created in the district by rumours of ‘evil-doers’¹¹⁵ who were using spells to kill babies in their cradles in such a manner as to resemble natural death. The murderers then dug up the little corpses and ate them, using some of the flesh to make an ointment that conferred magical powers, including flight and shape-shifting. The judge used torture to extract confessions from the accused, not just of baby-killing but of causing a range of injuries to people, as well as offering magical aid to customers. Some seem to have been solitary operators, but ‘Peter’ had no doubt that many belonged to a devil-worshipping sect with an initiation rite which included abjuring Christianity and drinking a liquid distilled from a dead infant. He burned both those who were made to confess and those who refused to do so, and seems to have claimed many victims.

For a long time, historians thought that this was the earliest datable reference to the new construct of the satanic witch cult, because Nider said that ‘Peter’ had conducted his campaign sixty years before. Joseph Hansen identified him in 1900 as a Peter von Greyerz who had governed the upper Simmen Valley in the 1390s, and this became generally accepted. Recently, however, the ‘Lausanne cluster’ has found two other judges called Peter who ruled the district between 1407 and 1417, while those in office between 1418 and 1424, and 1429 and 1434, are unknown, and could also have included a Peter. Moreover, von Greyerz had a son of the same name, active between 1421 and 1448, who would have been known to Nider, and who may have been confused by the friar with his father, just as reports of the Valais witch-hunt could have contaminated memories of what had happened in the Simmen Valley. Legal records survive from that valley which cover the years 1389–1415, and show no trials for witchcraft.¹¹⁶ The Valais hunt is therefore now the earliest based on the new stereotype of a satanic witch cult to be dated in the western Alps.¹¹⁷ What is significant about the Formicarius is that Nider unhesitatingly integrated all forms of magic, including the complex ceremonial kind and simpler spells sold to benefit clients, into that stereotype.

The next of the early texts from the region is the anonymous Errores gazariorum, ‘The Errors of the Cathars’, which was produced in the period 1435 to 1439 in two successive versions.¹¹⁸ Ponce Feugeyron has been suggested as a plausible candidate for its authorship, and another is George of Saluzzo, successively bishop of Aosta, to the south of the Valais, and of Lausanne: cases cited in the work were drawn from both dioceses. The name ‘Cathars’, coined for one of the most famous heresies of the high Middle Ages, was applied by this time to a range of outlawed sects, and was now given in the treatise to the imagined one of witches. This was defined as meeting in ‘synagogues’ (the word reflecting the contemporary suspicion and persecution of Jews as non-Christians) to worship the Devil, usually in animal form. The tract laid especial emphasis on the activities at these meetings, which included the eating of murdered babies (specifically under three years of age), dancing, and a sexual orgy without regard to the gender or kinship ties of partners. Initiates were given a box of ointment, consisting in part of baby fat, and a stick to anoint with it, on which they might ride easily to the ‘synagogue’. They also received powders manufactured of similar horrific materials in order to kill people – inflicting waves of lethal epidemics on communities – or to render them impotent or infertile, or blight their farmlands. The children were murdered in their beds at night, as alleged in the Valais trials, and then, as was also claimed there, dug from their graves and taken to the ‘synagogue’. Members of the sect pretended to be devout Catholics in daily life, and to offer comfort to the parents whose children they had killed.

The final work of significance among these early texts is that by Claude Tholosan, a lay judge in the French district of Dauphiné, who between 1426 and 1448 conducted 258 trials in the Alpine areas of that province and neighbouring parts of Piedmont, a region lying to the south and south-west of Lake Leman.¹¹⁹ His book, like that by Nider, grouped together all kinds of magical practitioner as members of the new sect of satanic witches, called by him ‘magicians’ or ‘evil-doers’. His portrait of the sect matches that in Fründ, Nider and the Errores gazariorum, save that he did not believe in the reality of its members’ ability to fly by night and he incorporated a theme from the folk tradition of the ‘good ladies’: that demons led initiates into wealthy homes to feast and make merry there, magically restoring the food and drink that they consumed. Records of trials conducted by Tholosan survive, and match the evidence provided in his book. Other trial records are preserved from the districts north of Lake Leman between 1438 and 1464, and show the imposition by clerical inquisitors, most partnered with Bishop George of Saluzzo, of the portrait of the witches’ ‘synagogue’; especially that provided in the Errores gazariorum which had drawn in turn on the earliest surviving trials in that region for its second edition. Those accused by other local people were arrested and systematically threatened, cajoled and tortured until they confessed to engagement in the list of activities attributed to the sect and named other members of it. Most were then burned, men forming a slight majority of the victims.¹²⁰

The literary works and the trials therefore had an interdependent relationship, drawing on a compact set of territories in the western Alps, which was – as Hansen long ago pointed out – to play a decisive role in engendering the early modern witch trials as the literary works produced there developed between them the portrait of what became widely known in the next century as the witches’ sabbath. These Alpine sources bear out Grimm’s model of a mixture of orthodox Christian portraits of heresy and folklore: but the two are not evenly balanced. The basic framework of the new image of satanic witchcraft in the two kinds of Alpine source was taken from images of heresy: the nocturnal gathering to worship Satan or one of his lesser demons, often in animal form; the indiscriminate sexual orgy; and child-murder and cannibalism; all adding up to an incarnation of the anti-human, derived from ancient times, as well as the anti-Christian. What the new stereotype did was to combine the Church’s two prior stereotypes of demonic heretics and demonic magicians, in an atmosphere ultimately produced by the persecution of magical practitioners, which commenced in the 1370s. The true emotive power of the combination was that it produced an heretical sect in which Satan empowered its members with the ability to work harm against their neighbours on a grand scale, with the aid of devils; and above all to kill their small children. This provided the context in which local panics could occur which led immediately – as seen in the Valais – to trials and executions on a scale out of all proportion to those of medieval magicians hitherto. The western Alps matter because they generated the texts that were to propagate this new concept of heresy; but as seen, they were but one district in a much wider region over which it had been hunted in the 1420s.

Carlo Ginzburg has argued (against Cohn) that older stereotypes of heresy were less important in the formulation of that of satanic witchcraft than specific persecutions of lepers and Jews in the mid-fourteenth century.¹²¹ It is true that the most serious accusation against Jews, of spreading plague, arose at a popular level in the western Alps, where the first texts concerning the witches’ sabbath were produced less than a century later. He is also correct that whereas the other constituent parts of older images of heresy continued in action during the fourteenth-century persecutions, child-murder and cannibalism did not. On the other hand, the lepers and Jews attacked in that century were not accused of most of the practices attributed to the new-style witches; but heretics such as the Waldensians were. The reintroduction of child-killing to the model, which it was suggested here was carried out by a campaign of preaching friars in the early fifteenth century, may have embodied ideas derived from older works of literature, or may have been a fresh start sparked by specific local anxieties. Ginzburg also, as has often been stated in the present book, emphasized the folkloric contributions to the new model, and especially those of night flight and shape-shifting to animals. However, the association between demons and animal forms was long established, as shall be discussed in a later chapter of this book. Michael Bailey has questioned Ginzburg’s stress on night flight, pointing out correctly that not all of the early texts mention it.¹²² It is not recorded in the Aneu Valley either; but it is at Todi, and is cited sufficiently in the early trials in western Switzerland to give it real importance. These citations, however, do not emphasize the spontaneous flight, or riding on animals, of the processions of the ‘good ladies’. Instead an ointment is portrayed as vital to the process, applied to the body at Todi or, in the Alps, to a chair or stick; which is not a motif found in the accounts of followers of the ‘ladies’. It is a much more ancient one, associated with the figure of the strix, and of the Roman witch. This had persisted through the Middle Ages, as proved by a Tyrolese poet of the mid-thirteenth century who mocked those who feared female cannibal witches who flew to attack people by night on a calf-skin, broom or distaff.¹²³It is tempting to relate the animated stick to the staff on which some magicians ride in the early medieval Norse literature, but the distance in time and space may be too large to make this tenable. It may therefore be proposed that Norman Cohn’s model of the origin of the early modern witch figure (which was itself an updating of Jacob Grimm’s) remains essentially correct, and may be restated now with better evidence and greater detail. Ideas derived from official notions of heresy were most important in the construction of the new belief system, and the most probable underlying ancient myth is that of the Mediterranean child-killing night-demoness, blending in the Alps into that of the Germanic cannibal witch. That belief system might, however, still have proved a short-lived phenomenon had not the Alpine hunts produced a body of texts to codify and promote it. As is well known to historians, fortune provided a perfect vehicle for this work, in that a major church council met at Basel, on the fringe of the western Alps, between 1431 and 1449 and represented for that time the major point for the development and exchange of ideas in Western Christendom. Nider, Feugeyron and George of Saluzzo were all present, and the Formicarius and Errores gazariorum seem to have been written there.¹²⁴ Others who attended the council, from outside the Alpine region, subsequently wrote works of their own to propagate belief in the new satanic conspiracy of witches, and these formed part of an extensive body of publication by French, Italian, Spanish and German authors which debated the reality of the conspiracy, and, increasingly, supported the idea of it.¹²⁵ Occasionally, it is possible to see in them once more the process by which existing local practices of magic were sucked up into the stereotype. One clear example of this is in the work of Pierre Marmoris, a professor at the University of Poitiers, who wrote in the early 1460s. He was as yet bereft of examples of the new satanic witch cult in his part of western France, and so he cobbled together cases of local magic which he had himself encountered, as examples of the menace from witches: people whom he had seen speaking incantations to heal animal bites or scare crows off crops; a man of whom he had heard at Chalons sur Marne who could make himself invisible; a Poitiers woman he had exorcized who claimed to be bound by an erotic spell; a Bourges man who offered to teach him how to refine wine at a distance; and legal prosecutions of which he heard for magic to cause impotence, and for the use of the hand of a corpse to put men to sleep.¹²⁶ From such trivialities a portrait of a major new satanic cult could be fabricated. The spread of trials of people alleged to belong to it, across parts of France, Italy, Germany and the Netherlands during the rest of the fifteenth century, showed the same ability to pick up pieces of local practice and lore. Likewise, although the basic stereotype of satanic witchcraft (secret assemblies to worship the Devil, followed by acts of destructive magic) remained constant, specific features of it – acts of homage to Satan, cannibalism, child-murder and orgiastic sex – were adopted selectively in these trials. Some had all these features, while others only a few, and not always the same selection, so that in practice a series of local variants was created, as they had been ever since the 1420s.¹²⁷ None the less, just as had been the case at that first appearance, a single basic concept was the driving force for local persecutions, and this was to remain more or less unchanged to produce the much more extensive trials of the sixteenth and seventeenth centuries.